AI-heavy private schools promised twice the learning; data shows a gap

The promise and the test

Alpha School markets itself on a striking claim: that its AI-driven approach lets children learn roughly twice as fast as in conventional classrooms. A New York Times examination set that claim against the network's own data and found that around half of its kindergarten-to-eighth-grade students did not hit the doubled-pace target in math.

Why parents are watching

The story lands in the middle of a larger experiment. Schools and families are being sold AI tutoring and AI-generated lesson plans as a fix for teacher shortages and stalled test scores, often with little independent evidence behind the marketing. Alpha's figures are among the few public data points, which is why they draw attention far beyond its own campuses.

What the numbers do not settle

A single metric in one subject cannot show whether AI instruction helps or hurts. Supporters argue that software which tracks mastery lets students move at their own speed; critics worry about screen time, thin supervision and inflated expectations. For parents weighing an AI-heavy classroom, the practical questions are what a school measures, how often it publishes the results, and whether a teacher or an algorithm decides what a child does next.